The Impact of Non-uniform Electrodes on Performance and ... · © 2016 Johnson Matthey Fuel Cells...

© 2016 Johnson Matthey Fuel Cells Ltd

The Impact of Non-uniform Electrodes on Performance and Gas Cross-over

Ed Wright, Emily Price, Jonathan Sharman

Johnson Matthey Technology Centre


Challenges for CCMs in PEMWE

Main target for PEMWE is cost reduction

• CCM one of the main focusses of cost reduction

• Thrifting PGM content in catalyst layers

• Novel catalysts both supported and unsupported being developed

• Reducing membrane resistance

• Thinner membranes

• Reducing hydrogen crossover

• Modifying chemistry / adding recombination catalyst

• Need to do above while maintaining performance and durability

• Porous titanium current collectors also add a significant cost

• Tighter tolerances significantly increase costs

• CCM needs to interact well with the current collector to allow reduction in hardware costs


Cell sealing of typical single cell

A mismatch between seal height and MEA step height may lead to compression issues

Membrane expansion on heating and hydrating may affect sealing

Ti Plate

Steel Plate

Seal Ti Sinter

Connection point to power supply

• Cell compressed by 9 bolts on edge, using fixed torque

• Assembly carried out at room temperature and with dry MEA

• Heating cell and hydrating MEA will lead to expansion of different components and so may affect compression


Sinter thickness

All sinters nominally 1 mm thick

GKN sinters have 40 µm variation within a sheet and 90 µm across the batch

Mott sinters have <10 µm variation within a sheet and ~10 µm across the batch

Weight / Density shows similar trends

4

0.8

0.9

1

1.1

1.2

1.3

1.4

GKN Mott 1-9 Mott 10-17 Mott 18-25

Sin

ter

Thic

kne

ss /

mm

© 2016 Johnson Matthey Fuel Cells Ltd © 2016 Johnson Matthey Fuel Cells Ltd

20.5 21 21.5 22 22.5 23 23.5 24 24.5 25 25.5

-100

-50

0

50

Heig

ht

/

m

GKN

20.5 21 21.5 22 22.5 23 23.5 24 24.5 25 25.5

0

50

100

Distance / mm

Heig

ht

/

m

Mott

0 5 10 15 20 25 30 35 40 45 50-200

-100

0

100

Heig

ht

/

m

GKN

0 5 10 15 20 25 30 35 40 45 50-200

-100

0

100

200

Distance / mm

Heig

ht

/

m

Mott

Surface profiles of typical sinters

• Preparation method can have a significant effect on sinter

• Cutting method can affect long range shape (bowing) • Both sinters have similar roughness values

• Ra values of 10.9 and 10.2 for GKN and Mott respectively

• Gravity sintered (Mott) gives similar roughness above and below surface • Pressure sintered (GKN) has flat top surface with deep pores

5


Complete anode plate assembly

0 100 200 300 400 500 600 700 800 900-600

-500

-400

-300

-200

-100

0

100

200

distance / mm

heig

ht

/

m

GKN2c

Mott2a

Mott2b

Mott2c

Mott2d

Seal

Ti Plate Flow field / sinter Ti Plate

Seal

• Step height can vary up to 400 µm from plate depending on sinter and orientation

• Hard for < 200 µm CCM to accommodate such variation • What are the effects of the non-uniform compression?

Approximate CCM thickness


Experimental setup

• Baltic fuel cell QCF25 cell hardware used

• 5 x 5 cm active area

• Ti flowfield on anode and C flowfield on cathode • Ti sinter as anode current collector

• SGL 10BB carbon paper as cathode current collector (420 µm thick)

• Compression measurement device fitted • Piston to control clamping force on active area

• Linear transducer for displacement monitoring

• Current mapping included • S++ system fitted behind cathode flowfield

• 100 segments measured

• In house test station operating at 60 °C, ambient pressure, 500 ml min-1 water flow rate

• Hydrogen crossover measured with TCD

• Tests carried out with in-house MEAs


Effect of increased compression

• MEA: IrO2 (2mg cm-2) | N117 | Pt black (1 mg cm-2)

• Operation at 500 mA cm-2 (12.5 A total)

• Compression increased from 0.5 – 5.5 bar (0.16 – 1.72 N mm-2) in 0.5 bar steps

• Cell voltage drops by ~ 100 mV at 0.5 A cm-2 operating point as compression increased

• Displacement of 60 µm occurs as clamping force increased – likely to be carbon paper compressing

8


Effect of increased compression - crossover

• Crossover decreases as compression increased

• All other cell parameters kept constant

• Crossover not expected to be affected by compression / operating voltage

9


Current mapping during compression

• Compression stepped from 0.5 bar to 5.5 bar in 10 steps of 120s

• Histograms show number of segments (%) operating within a current range

• Far left bar indicates number of inactive segments

• At low compression 70% is inactive, at 2.75 bar 30% is inactive and at 5.5 bar ~ 10% is inactive

10

0.5 bar

2.75 bar 5.5bar

5.5bar


Segmented polarisation curve 2 bar compression

50 A

15.6 A

• Uncoated Mott sinter used with 10BB cathode GDL

• Test carried out after compression test so GDL may be compression set

• Significant performance difference between segments

• Some segments are saturated so not reporting true current density

• ~30% of area inactive

Time / s

Cu

rre

nt

De

nsi

ty /

A c

m-2


Segmented polarisation curve 5.5 bar

12

50 A

15.6 A

• Significant improvement in uniformity

• Still significant region inactive (15 – 20%)

• Fresh GDL may improve uniformity further

Time / s

Cu

rre

nt

De

nsi

ty /

A c

m-2


Overall Polarisation

• Crossover can be seen to increase as current density increases

• Typically expect fixed diffusion through membrane and so lower crossover at higher currents (increased O2 production)

• Effect more dramatic for lower compression / less uniform layer

0 0.5 1 1.5 20.5

1

1.5

2

2.5

3

Current Density / A cm-2

Low Compression Voltage / V

High Compression Voltage / V

Low Compression Crossover / %

High Compression Crossover / %


Performance of supported Ir catalysts

• Layers with lower catalyst loadings or novel OER catalysts can show increased sensitivity to compression / contact resistance

• Effects cannot always be removed by increasing cell compression

• Over compression can crush carbon papers or crack plates

14

Data provided by SINTEF


Effect of sinter on catalyst layer

• Larger gaps in sinter may lead to inactive regions

• Effect will be exaggerated for thin layers • Effect will be exaggerated for low

conductivity layers • Low activity regions may experience

high voltages increasing oxidation of sinter and so rapidly deactivating completely


Effect on low loading layers

• Lower loading catalyst layers more sensitive to cathode catalyst layer

• Pt on C cathode will be thicker and more compressible than Pt black

• Conventional IrO2 layer printed at ~ 2 mgIr cm-2

• Novel IrO2 layers printed at ~ 1 mgIr cm-2

• Thin reinforced membrane with recombination catalyst used for all samples

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 20

0.5

1

1.5

2

2.5

3

Current Density / A cm-2

IrO2|Pt blk Voltage / V

novel IrO2|Pt blk Voltage / V

novel IrO2|Pt on C Voltage / V

IrO2|Pt blk Crossover / %

novel IrO2|Pt blk Crossover / %

novel IrO2|Pt on C Crossover / %


Current Mapping

Novel IrO2 1 mgIr cm-2 anode, Pt/C cathode, recom cat membrane, 5.5bar,

60oC

Novel IrO2 ~ 1 mgIr cm-2 anode, Pt black cathode, recom cat membrane, 5.5bar,

60oC

IrO2 black 2 mgIr cm-2 anode, Pt black cathode, recom cat membrane, 5.5bar,

60oC

• Current maps taken from highest current density

• Conventional 2 mgIr cm-2 layer shows most uniform activity

• Novel 1 mgIr cm-2 anode with Pt black cathode shows worst uniformity

• Novel 1 mgIr cm-2 anode with Pt on C cathode more uniform but not as good as thicker commercial IrO2 layer

• Cathode / overall CCM thickness may be important for good contact


Conclusions

• Poor compression leads to non-uniform PEMWE performance

• Gas crossover increases with decreasing uniformity

• Crossover seen to increase at high current densities for non-uniform layers

• Sinters vary considerably in thickness, curvature and contact with CCM

• Thickness variations of similar magnitude to total CCM thickness

• Poor compression / contact leads to non-uniform performance

• Novel OER catalysts or reduced loadings will exaggerate the effects

• Cathode type / thickness can help improve uniformity / crossover

• Thinner membranes will reduce the ability of the CCM to compensate for hardware variations

• Machining components to give < 100 µm variation for plate / current collector combination not practical

• CCM / Hardware interactions need to be well understood to help reduce costs


Acknowledgements

• NOVEL Project partners particularly

• Magnus Thomassen, Tommy Mokkelbost and Alejandro Oyarce Barnett at SINTEF

• Tom Smolinka and Thomas Lickert Fraunhofer ISE

The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013) for the Fuel Cells and Hydrogen Joint Technology Initiative under grant agreement n° [303484] (Novel)

Date post:	06-Aug-2020
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

The Impact of Non-uniform Electrodes on Performance and ... · © 2016 Johnson Matthey Fuel Cells...

Documents